ABSTACT

Objective

To assess the manifestations of pain in PD (Parkinson’s disease ) patients versus healthy controls.

Methods

Data on pain was collected from 127 patients and an equivalent number of controls using two self-report questionnaires: the Brief Pain Inventory (BPI) and a custom-made questionnaire focusing on specific details not covered by the former. Additional analysis was conducted within the patient group only to analyze the potential effects of factors relating to PD on the various measure of interest relating to pain.

Results

Parkinson disease patients had lower odds of experiencing pain in both arms (ExpB=0.061, p<0.001), greater probability of demonstrating pain in both legs (ExpB=2.409, p=0.024), and an increase difficulty in localizing pain (ExpB=2.958, p=0.030). There was no relationship between duration of pain (F=12.414, p=0.001) or arthritis (ExpB=0.724, p=0.309) and pain in PD. The likelihood of experiencing nagging pain (ExpB=3.533, p=0.028), but not other forms, was much more strongly associated with PD patients than normal controls. When all other types of pain were controlled for, pain in PD is more likely associated with akathesic pain (ExpB=9.046, p<0.001). Conclusion

There are major differences between pain in PD patients and pain in normal controls, which could have implications on the pathophysiology and adequate management of pain in different populations.

Keywords: Healthy controls, movement disorders, pain, Parkinson’s disease

INTRODUCTION

Description and classification of pain in PD

While Parkinson’s disease (PD) classically presents with tremor, rigidity, akinesia/bradykinesia and postural instability, patients often experience a much broader scope of symptomology [4, 27]. Later stages of this disease may manifest as dementia, depression, or psychosis [22]. Furthermore, over 90% of PD patients may experience non-motor symptoms, including pain [19]. Several studies, such as the case study by Djaldetti et al., have demonstrated that pain is much more frequent in patients with PD than in the healthy population [9]. Roughly 30-85% of patients with PD complain of experiencing pain, the large variation likely being accounted for by varying definitions and assessment tools [2]. Pain can be classified as either primary or secondary, the latter being more prevalent. Further classifications are achieved on the basis of major symptom etiologies [24].

Primary pain, or pain of central origin, is defined as pain that arises directly from the disease process itself [2, 24]. Patients may present with general discomfort or malaise. Alternatively, they may experience unexplained sensations in unexpected locations. These may be painful, stabbing, itching, burning, aching or tingling in character [24].

On the other hand, secondary pain arises from states that are consequent to PD [17]. The major etiologies of secondary pain include radicular/neuropathic pain, musculoskeletal (MSK) pain, akathisia, pain related to restless legs syndrome (RLS), and dystonia-related pain [24]. Radicular/neuropathic pain is due to a direct insult of the nervous system itself, and is characterized by persistent tingling, shooting, electrical or parasthetic sensation [1]. Pathology of the muscular and skeletal structures of patients with PD may lead to MSK pain. This is a potential consequence of rigidity and akinesia, which are both common early features of PD [16]. Akathisia is described as a psychological and physical irritability, where the patient is unable to remain still. It is a potential side effect of prolonged exposure to typical antipsychotic medications, such as Haloperidol, but may also be present in PD patients [24]. This is in contrast to RLS, where patients feel the urge to move but maintain voluntary control over their muscles. Furthermore, only RLS is tied to circadian rhythms [3]. Restless legs syndrome is diagnosed by the presence of the following four features: the urge to move limbs, accompanied by unpleasant sensations and feelings; worsened by rest and inactivity; relief through getting-up and moving around; and urges being the worst in the evening [3]. Of all the secondary causes of pain, dystonia is the most intense. Dystonia is a movement disorder characterized by painful twisting and unusual postures, which are secondary to sustained muscle contractions [15].

Management of pain in PD

Management of pain in PD may be achieved through either pharmacological or non-pharmacological approaches, or a combination of both [2]. The classification of pain as either primary or secondary helps to determine the ideal treatment regime. The sub-classification of secondary pain plays an additional role [24]. For example, the optimal treatment for MSK pain is a combination of MSK examinations; physical/occupational therapy; medical, anti-inflammatory, and dopaminergic therapies; and surgical alternatives, such as decompressive surgery when indicated [24]. On the other hand, pain related to akathisia is best managed with dopaminergic therapy, opioids and benzodiazepines [24]. Thus, safe and effective treatment strategies do exist. However, more focused research is needed to maximize effective pain management in PD patients [24].

Direction of our study

Pain affects individuals both physically and psychologically, ultimately impairing their ability to perform daily activities. Despite such a high prevalence, pain has yet to be well studied among PD patients. In our current study, we compare the prevalence of pain and its potential causes among PD patients to a similar group of control subjects. Among the individuals who experience pain, the reported severity, sensation, duration, and location are also analyzed.

MATERIALS AND METHODS

Study participants

Data was collected from 127 patients and an equivalent number of controls by a neurologist at a community-based Parkinson’s disease and movement disorders center in 2013. Patients were diagnosed with idiopathic PD though UK Brain Bank Criteria, and were regularly assessed three to four times a year for follow-up care [14]. Patients ranged in age from 37 to 93 years. All patients were informed about the nature of the study and gave their written consent for participation. The local ethics review board approved of the study.

Study methods

Two self-report questionnaires were used: the Brief Pain Inventory (BPI) and a custom-made questionnaire focusing on specific details not covered by the former, including items specific to PD [6]. Participants were screened based on the presence of focal or generalized body pain, such as headaches. A negative finding would exclude the patient from our study. Participants provided information for potential causes by giving accounts of past medical history as well as other relevant events. Medical records were available to provide verification and additional details. The variables of interest relating to pain were compared within and between the groups to identify any differences in manifestation and/or cause. Patients were asked details on the nature of the pain experienced, such as onset, location, alleviating/exacerbating factors, quality, radiation, severity and timing. Additional sequelae, such as tingling, burning or cramping were assessed. Further analysis was conducted within the patient group to analyze the potential effects of factors relating to PD (side of body affected, Hoehn and Yahr stage, UPDRS score, etc.) on the various measure of interest relating to pain [13, 18]. Exclusion criteria included patients with atypical parkinsonism, drug-induced parkinsonism, as well as pain due to external causes and/or cognitive dysfunction.

Statistical methods

The causes and manifestations of pain in PD patients versus healthy controls were compared and assessed with Pearson correlations and multiple logistical regressions using SPSS version 19 (SPSS Inc., Chicago, IL, USA). A significance level of 0.05 was used throughout all of the analyses.

RESULTS

Pain in 127 patients with PD was compared to pain in an equivalent number of controls. Patients were assessed for the location of and ability to locate pain; the relationship between arthritis and duration of pain to pain in PD patients versus controls; the likelihood of experiencing various qualities of pain; as well as differences in the specific types of pain experienced.

Firstly, it was found that patients with PD had a decreased likelihood of experiencing pain in both arms when compared to controls (ExpB=0.061, P<0.001, Table 1). Furthermore, pain in either arm was not significantly correlated to the presence of PD in pain patients. Conversely, the odds of experiencing pain in both legs increased 2.4-fold in patients with PD (ExpB=2.409, P=0.024, Table 1). Interestingly, pain is almost three-times as difficult to locate in patients with PD when compared to normal controls (ExpB=2.958, P=0.030, Table 1). Further analysis demonstrated that there was not a significantly increased propensity of experiencing pain in PD for patients with arthritis (ExpB=0.724, P=0.309, Table 2) or a longer duration of pain (F=12.414, P=0.001, Table 3). When assessed for likelihood of experiencing certain qualities of pain in PD patients versus healthy controls, only nagging pain was reported as being significantly more likely. Indeed, there was a 3.5-times greater probability of experiencing nagging pain in subjects with the onset of PD as opposed to those without (ExpB=3.533, P=0.028, Table 4). There was no difference in likelihood for aching, dull, tension, sharp, penetrating, shooting, throbbing, gnawing and numb pain. Radiating, boring, burning, stabbing and tender pain were excluded due to low sample size. When subjects were assessed for the likelihood of experiencing various types of pain in PD versus non-PD, patients with PD were found to be 9-times more likely to experience akathesic pain when other types of pain were held constant (ExpB=9.046, P<0.001, Table 5). Other forms of pain include non-localized, dystonic, and neuropathic pain. When akathesic pain was not controlled for, neuropathic pain became significant in PD patients (ExpB=2.530, P=0.018, Table 5). Further analysis found that patients with pain and PD had an increased chance of experiencing an interaction between akathesic and neuropathic pain (ExpB=3.862, P=0.027, Table 5). A summary of the PD+ and PD- populations that experienced certain locations, qualities, and types of pain is highlighted on Table 6. DISCUSSION

Location of pain

Pain in PD is well recognized. Three recent reviews highlight the wide range of pain symptoms associated with PD and describe new ways of characterizing and classifying nociceptive, neuropathic and musculoskeletal pain of Parkinson’s disease [10, 12, 29]. Our study found that PD patients are more likely to experience pain in both legs as compared to arms. Limb pain in the form of unilateral or bilateral shoulder pain, hand and leg pain has previously been described in relation to PD and a recent review summarizes these and other locations of pain [10, 26]. Another recent study describes persistent leg pain as a distinct entity and provides information specific to this presentation [28]. This may help explain why PD patients are more likely to experience leg pain than arm pain.

Localization of pain

PD patients experiencing pain had greater difficulty identifying the precise location of pain when compared to healthy individuals. This may represent a novel finding, as to our knowledge, there is no literature comparing PD patients ability to localize pain to healthy individuals. We also found that PD patients were more likely to report their pain as aching when compared to healthy individuals. This is in agreement with another recent study, which found that the most common physical experience of pain in PD patients was aching [21].

Quality of pain

Our study demonstrates that patients with PD are 9-times more likely to experience akathesic pain when compared to healthy controls. It has already been established that Parkinson’s disease is closely associated with akathisia, as both seem to arise from the basal ganglia [25, 30]. Indeed, akathisia has been found to occur in up to 45% of PD patients, and correlates with both the age of onset and severity of PD [7]. However, the basis of this association remains controversial, and has been postulated as a consequence of dopamine receptor blocking drugs, and not PD itself [21]. Another theory is that alterations in the mesocortical pathway, which is associated with the cognitive defects of PD, may be involved in the pathophysiology of akathisia [20]. Although a clear etiology has yet to be defined, our study adds further support to the increased prevalence of akathesic pain in patients with PD, and may provide insight for how patients can better manage their pain.

MSK comorbidities and pain in PD

The fact that patients with PD and arthritis were not more likely to experience pain than those with PD alone removes arthritis as a comorbidity that may compound our findings. Musculoskeletal problems are very common in PD as well as with the general population, and an important contributor to a decreased quality of life [31]. However, studies on the prevalence of arthritis and PD are conflicting, with some showing an increased prevalence, and others not [8, 23]. If there is in fact a greater prevalence of MSK problems in PD patients, the management of these symptoms should not be neglected, especially in post-deep brain stimulation (DBS) patients where it is a common complaint [11]. In addition, past studies have shown a positive correlation between MSK problems and pain in PD patients [31, 32]. Despite this possible association, pain in PD patients within our study population seems to have not been affected by arthritis. One possible explanation is that the MSK pain was overpowered by a concomitant akathesic pain. Regardless of the reason, more studies should be performed to clarify this discrepancy, and to elucidate a more definitive relation between arthritis and pain in PD.

Strengths and weaknesses of study

Compared to studies reported in the literature, the present study had a relatively large sample size and adopted a novel approach to the study of pain in the PD population. This approach, in addition to the standardized assessment methods used which may limit biases, is a significant strength of this study and was shown to be applicable to the study of PD-associated pain. Importantly, research with questionnaires filled out by PD patients and their caregivers may more accurately reflect feelings of pain in the life of PD patients. Excluding patients with cognitive dysfunction and dementia also helped to ensure the validity of patient accounts throughout the interview process. Another strength was that our analysis accounted for the effect of various localizations and types of pain. Our study is not without limitations, however. Neurophysiologic examinations may have been useful to objectify the classification into pain types in some cases (e.g. better define what is meant by “aching pain”). Secondly, the subjective nature of pain leaves some limitations to all studies of pain in humans. Finally, our group of healthy individuals was not aged-matched controls, thereby possibly limiting the strength of our results.

CONCLUSIONS

In spite of pain being a common NMS in PD, it continues to be under-reported and under treated as demonstrated in a recent study across European centres [5]. It is thought that these symptoms may remain undeclared to health-care professionals due to a lack of awareness by the patient that the symptoms are linked to PD. This highlights the need for enhanced awareness and a better understanding of pain in PD by clinicians. We therefore feel that the publication of more detailed information regarding pain phenotypes in PD is important, as this will promote early recognition and management of PD related pain. Our long-term goal would be to increase the chance of effective management, enabling a better quality of life for patients affected by this debilitating disease.

Table 1. Location of and ability to pinpoint pain in Parkinson’s disease patients versus healthy controls.

Table 2. The relationship between arthritis and pain in Parkinson’s disease patients.

Table 3. The relationship between duration of pain (hours) and pain in Parkinson’s disease patients.

Table 4. Likelihood of experiencing various qualities of pain in Parkinson’s disease patients versus normal controls. Radiating, boring, burning, stabbing and tender pain were excluded due to lack of sample size.

Table 5. Likelihood of experiencing various types of pain in Parkinson’s disease patients versus healthy controls. Pain type likelihood is assessed when all forms of pain are held constant and when all types minus akathesic pain are controlled for. Interactions between akathesic and neuropathic pain are also analyzed.

Table 6. Summary of the population that experienced pain with (PD+1) and without (PD+0) Parkinson’s disease.